| 基于纹理特征的高分辨率遥感图像土地利用分类研究 |
| |
| 引用本文： | 胡玉福,邓良基,匡先辉,王鹏,何莎,熊玲.基于纹理特征的高分辨率遥感图像土地利用分类研究[J].地理与地理信息科学,2011,27(5). |
| |
| 作者姓名： | 胡玉福 邓良基 匡先辉 王鹏 何莎 熊玲 |
| |
| 作者单位： | 四川农业大学资源环境学院,四川雅安,625014 |
| |
| 基金项目： | 国家星火计划资助项目(2005EA810087); 四川省科技支撑计划项目(2009NZ0051) |
| |
| 摘 要： | 为了提高高分辨率遥感图像土地利用分类精度,该文以金沙江下游河谷地带SPOT 5遥感图像350×350像元作为试验区,在ERDAS IMAGINE 9.0和ENVI 4.1软件平台支持下,采用灰度共生矩阵方法提取遥感图像对比度、角二阶矩、熵、同质度等纹理指标辅助遥感图像分类,分析结果表明,相对于传统监督分类方法,基于纹理特征辅助监督分类方法总Kappa系数提高了9.15%,耕地、林地、水域、建设用地、未利用地Kappa系数分别提高了7.60%、6.17%、3.59%、15.74%和2.96%,分类结果面积准确率分别提高了3.38%、13.47%、4.65%、10.22%和1.53%,说明纹理辅助监督分类方法相对于传统监督分类方法有效提高了土地利用分类精度。
|
| 关 键 词： | 遥感 纹理分析 灰度共生矩阵 监督分类 |
Study on Land Use Classification of High Resolution Remote Sensing Image Based on Texture Feature |
| |
| HU Yu-fu,DENG Liang-ji,KUANG Xian-hui,WANG Peng,HE Sha,XIONG Ling.Study on Land Use Classification of High Resolution Remote Sensing Image Based on Texture Feature[J].Geography and Geo-Information Science,2011,27(5). |
| |
| Authors: | HU Yu-fu DENG Liang-ji KUANG Xian-hui WANG Peng HE Sha XIONG Ling |
| |
| Institution: | HU Yu-fu,DENG Liang-ji,KUANG Xian-hui,WANG Peng,HE Sha,XIONG Ling(College of Resources and Environment of Sichuan Agricultural University,Yaan 625014,China) |
| |
| Abstract: | In order to improve land use classification accuracy of high resolution remote sensing images,in this paper,in support of software of ERDAS IMAGINE 9.0 and ENVI 4.1,some texture indexes of SPOT5 image of downstream region of Jinsha River,such as contrast,angular second moment,entropy and homogeneity had been extracted using the gray level co-occurrence matrix.And these texture indexes had been used to assist supervised classification.The results showed that,compared with the method of traditional supervised... |
| |
| Keywords: | remote sensing texture analysis gray level co-occurrence matrix supervised classification |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
